Full framework
    gem

    Layer L5

    Execution

    Applied skills and capabilities. Doing the actual work.

    Why it matters

    A jeweler takes refined gold and crafts rings, necklaces, watches, each requiring specialized skill. L5 is THE entry point for AI-native companies.

    The Master Jeweler

    A jeweler takes refined gold and crafts rings, necklaces, watches, each requiring specialized skill. In AI: domain skills, decision frameworks, and operating playbooks transform generic intelligence into specific capability. Harvey knows legal. Sierra knows CX.

    The 5 sublayers

    L5a

    Domain Execution & Tool Use

    Doing the actual work, legal drafting, code generation, diagnosis, underwriting, including function calling, code interpreter, browser/computer use, and structured tool invocation that turns a model into an operator

    L5b

    Decision Frameworks & Reasoning Scaffolds

    Structured thinking patterns, checklists, rubrics the agent follows

    L5c

    Retrieval-Augmented Workflows

    Grounding execution in retrieved context, knowledge, and documents

    L5d

    Operating Playbooks

    Company-specific SOPs, rules, preferences encoded for agents

    L5e

    Interaction Skills & Actuation

    Tone, empathy, negotiation, persuasion, and physical-world actuation (robotic control, valve/vehicle/device operation)

    , Layer diagnostic card · SCOI v1

    Is a company really at L5?

    Applied domain skill, the layer that actually performs the economic work (legal drafting, code, diagnosis, CX resolution).

    Inclusion tests · include if ALL

    • Performs a specific economic task end-to-end, not just generates text about it.
    • Embeds non-obvious domain decisions (L5a/b/d), rubrics, SOPs, playbooks a generic L2 wouldn't know.
    • Buyer pays per outcome or per workflow, not per token.

    Exclusion tests · exclude if ANY

    • A prompt template dressed as a product, generic L2 reasoning, no embedded domain logic.
    • Generates artifacts but humans still do the actual decision work.
    • Calls itself 'AI for X' but L2 alone matches its output quality.

    The L5 removal test

    Swap the L5 layer for raw L2 + a clever prompt. If the output is indistinguishable, there is no L5, only an L7 with vertical paint.

    Economic work this layer does

    Encodes the part of expertise that is *not* in the training corpus: decisions, exceptions, escalations, taste.

    Canonical examples

    • Harvey

      L5 (legal-specific reasoning + drafting) tied to L1 corpus + L3 trust. Fortress.

    • Sierra

      L5 CX resolution + L8 per-tenant memory. Outcome-priced, not seat-priced.

    • Cursor

      L5 code-editing skill + L4 IDE surface + L8 repo memory.

    Anti-examples · look-alikes that fail

    • Devin (Cognition)

      L5+L7 pitched as 'an agent', thin on L1/L8, easily out-shipped.

    • Generic 'AI SDR' tools

      L7 sequence-writer with no L1 behavioral data or L8 account memory.

    • Most "AI for [vertical]" decks

      Prompt + vertical landing page. L7 with L5 marketing.

    Disagree with a classification?Open the classification table →

    Who's playing here

    HarveySierra11xCursor

    Verdict: Durable if deep. Generic skills get absorbed.

    Case studies touching L5

    Sierra's Memory Moat: Why L8 Beats Salesforce's Agentforce

    Sierra and Salesforce Agentforce look like the same product on stage, an AI agent that resolves customer issues. The Cube projection shows they are structurally opposite. Sierra was architected as L1c behavioral data + L5d operating playbooks + L8c network learning from day one: every resolution compounds into per-customer memory. Agentforce is L5 bolted onto Salesforce's existing L1, with no compounding loop. Same demo, opposite trajectories.

    From Dashboard to Skill Hire: The Death of Per-Seat Software

    Software has moved through five distinct eras of human–machine division of labor. We are mid-transition between Era 3 (The Dialogue, human directs, AI builds) and Era 4 (The Workspace, AI orchestrates, human supervises). Era 5 (The Skill Hire, the agent IS the worker) arrives by 2028. Per-seat pricing is structurally dead in Eras 4–5 because the seat itself goes away. Every product roadmap needs to be re-priced and re-architected along both the customer axis and the depth axis.

    Harvey AI Through the Layers

    Harvey is built across four sublayers, L1b (licensed case law), L3a (compliance gates), L5b (legal reasoning scaffolds), L8d (institutional memory of matters). A useful case for mapping how a vertical-AI company actually stacks up, and where horizontal platforms can and can't reach.

    Klarna: 700 Agents Replaced, $40M Saved, The First Honest Number on Agent Economics

    Klarna's AI assistant handled 2.3M conversations in its first month, the workload of 700 human agents, with equal customer satisfaction and faster resolution. The headline is the cost. The structural story is that Klarna owned the customer data (L1c), the workflow (L5a), and the resolution memory (L8c). The model was a commodity input.